Thinking About The "Music Web"

When I write about things like the music business I am generally talking out my ass and trying to make sense of how things currently work, and how artists could be done better. The thing I always try to do when I look at a new area of interest is find out how everyone else thinks about it, and then see if there’s another way to look at it. Right or wrong, that change in perspective at least gives me ideas for something to play with or make. My methods may make me look stupid at first, but frequently these methods also put me way far ahead of the curve.

If you take the state of things in the online music world, they are currently dominated by a few beliefs/events:

1. Napster being sued out of existence (and anyone who challenged the labels).
2. Websites permeating everything anyone wants to do these days.
3. The notion that music has to go through an intermediary for people to find it and enjoy it.

I’m not saying these are invalid or valid beliefs, they’re just the way things are, and I try to come up with alternatives. Who knows, maybe having the labels in control is really great. Maybe websites are the absolute best ways to do any application. Maybe having walled gardens and intermediaries between fans and the artists is the way to go.

However, I force myself to think about the situation with the assumption that everyone is wrong, and then I explore how it could be different. Many times the ideas are retarded, sometimes they’re great, but more importantly I develop an alternative perspective that makes me smile.

The danger in thinking about things differently (not right or wrong, just differently) is that “different” becomes wrong when presented to someone who thinks in the conventional way. When I take an idea I have about politics and present it to a Republican, they think I’m a Democrat. I say the same thing to a Democrat and I’m a Republican. I’m neither, I’m just different. I’ve got a third option that they haven’t considered, or that’s slightly different than both viewpoints, so their mind thinks that I must be a “them” not an “us”.

That’s perfectly fine by me. I enjoy being different, except for the annoying arrogance of the self-assured. It gives me fun ideas, and I just have to put up with an asshole or two along the way.

Yet, I’m Not So Different

What I found very interesting in various comments was that tons of other people are thinking about the exact same thing:

“Since this topic is right where my startup ( http://www.klictrack.com/mdt ) is trying to push things forward, please check us out and tell me where we need improvement. Any feedback would be greatly appreciated.” — klon (HN)

This wasn’t a rare comment, and given the number of sites I already mentioned, I predict there’s going to be a huge burst in the online music world very soon. One of these smart kids is gonna eat the lunch of the current intermediaries. Just watch. And no I’m affiliated with klictrack.com.

People who are currently making money as an intermediary in music, and scoff at what I’ve said so far on the subject should probably start paying attention. Not because I said it, but because that’s exactly how this story goes. C’mon we all know it. PanAm is the biggest airline in the world, and they think they’ll never die. Then a competitor comes along and makes running an airline more efficient, offers cheaper prices and bye-bye PanAm.

Go through companies like Dell and Wallmart and what you find is a bunch of companies that made their money finding inefficient chains of production and eliminating the middle men. Especially in a market with fixed prices, monopolies, and artificial scarcity like the music business.

All the people who think that needing a few middle men means that having tons of them is inevitable really need to go read up on history. Then again, maybe these guys are smart as hell and realize that it’ll just make them money for a short period of time.

Did You Notice My Trick?

In my previous blog post I said that the middle men aren’t really needed, and people scoffed. But, did you also read at the bottom where I pointed out that you could sell on Amazon anyway using their existing services:

• Flexible Payment System for payment processing.
• S3 for file serving and hosting.
• EC2 for software hosting.
• SimpleDB for data retention.

A few people got it right away, but most others just kind of skipped over that. I put important stuff like that at the bottom to catch the idiots sleeping at the wheel.

Right away someone sent me a link to Shrub which is a Google App Engine hosted S3 service that supports feeds for what you posted in RSS, JSON, and XSPF with the XSPF player already embedded. It’s totally open source, so if you’re one of the few people with a clue, then go grab it. Things could get interesting.

What about the payment? Oh, well here’s a Python lib that does FPS payments. I wonder if Shrub could have that too? Hmm.

Then Again, Maybe I Am Different

I really think someone could do a site, but man I am just bored with web sites. They don’t stretch my brain anymore, unless I were to try doing it in Haskell or something bizarre. Instead, I wanted to think about the problem from the point of view that current music publishing and consuming seems to go through these walled gardens. It’s like I’m stuck back in the day using CompuServe again.

Today I started just thinking about what a “music browser” would be, so I took a few notes of just random thoughts.

Reverse engineering the way the web works now, we have a few components to work with that make it function in its nice decentralized manner:

1. href takes the browser to another site without the originating site being involved apart from mentioning the href.
2. a document can have hrefs without having to register with the target.
3. documents have formatting and media.
4. anyone can setup a site and post their content as long as they follow a few simple protocols. Well, alright, reasonably less complex protocols.
5. the only control on publishing is the DNS system, which is fairly wide open and international.
6. consumers trust their browser to not snoop on them or do risky operations, so they don’t mind going to random sites. Whether that’s valid or not is debatable, but that’s the unwritten contract.
7. There’s an established strong standard for encrypting payment with SSL.

I’m sure there’s others but these came to mind as important for a music browser. Keep in mind that what I want is to get “eyeballs” to artists, but the way the web does it. I want people to go to a “site” and click around going to other sites. I don’t want them to start up iTunes and take what they’re fed.

In the above description, I think we can assume that things like HTTP, SSL, DNS, and browser trust can be taken for granted. I think I’ll just assume that whatever this browser does, it’ll use HTTP/SSL/DNS, and follow the same rules as a web browser. Maybe it could even be based on Mozilla or a plugin. Who knows, the point is, they work well enough for the task. Doing that then also tackles the hosting problem. You just host your music stuff for the browser on a web server like normal.

This also takes care of what goes in an href. I’ll just put URLs like normal in them, but down the line maybe we’ll just rename the starter to something mb:// so that a web browser will open the music browser when someone clicks it from an HTML page. That simplifies jumping people from a site to the music, and it’ll keep things straight for both the client and the server. The protocol wouldn’t change, just the name.

The Album Is The Document

That leaves only the problem of media, hrefs, and the document. When talking about sharing music with someone, what is the media? What becomes the document? Where do you put an href? What the hell does an href say? In HTML the document supports textual media and images, and then you can link that to another location via an href. How do I indicate in a piece of music that at 2:00 into the song, the user should be able to click a link to the artist I referenced for that killer riff? Would I put it in the actual music file, or in some other HTML-like document?

Let’s say that in our browser, the media is the music mp3 file, but the document is like a semantic HTML page someone can craft to talk about the music. It might reference artists I like, give my name, point at the blog, list pictures, other media, indicate that this song is part of a concept album, whatever. The document isn’t the song, it’s the information in computer readable form that talks about the music files.

In reality, a “document” would be the analog of an album in our world. You would write a document that described a cohesive work of art for your listener. Rather than disconnected songs on a site, you could step into the realm of creating a whole musical experience for someone by tying pieces of music, art work, links to related content, and textual information through a document.

With that distinction the hrefs are simple URLs inside the enclosing document, but then we’ve got the problem of time. Music hrefs would have to be timed to the music, and probably would need to work like VH-1 Pop-up Videos. Someone could see all the links, and just click them, or they could listen to the songs and watch them go buy, clicking if they like. I’m thinking some of the hrefs would need some kind of time code to indicate where they are relevant.

Finally, it might be interesting to add all the other features found in a browser. Javascript added to the document to create interactions? Something like CSS to style the music, maybe to indicate volume levels, effects, or possibly alternative versions of the same song.

The Music Web

This kind of exploration of the idea points me in the direction of a “music web”. Instead of going to a web site and just downloading disconnected sets of music, they could package sets of music as albums. The music “site” then becomes a collection of these album documents, and the albums contain references and media for the listener’s experience. However, since the listener is in control with a client that is on their computer, they can choose to present it however they want. Someone could write a Lynx version of this if they felt like it. Just like HTML they are in control of the presentation.

The hyperlinks from a document would then create a web of music that works the way our current WWW operates. As someone listens to a song, they could travel to other artists that are referenced. This is an interesting proposition because many sites try to sell consumers more music by guess what you like based on what you listen to currently. These sites rarely have the artists indicate who their inspirations are, let alone time these references to points in their music.

The very first major use I see of this isn’t finding music, but it’s remixing music. Imagine where you find a piece of music that someone built from another piece, and through their album/document you can find it, and automatically go grab it so you can get to the source. The music web could support hyper remixing of content, even going so far as to have music production software support the protocols directly allowing you to dump whole graphs of artist relationships into your song for exploration.

Problems (Obviously)

Right away I see the problem being that musicians probably couldn’t hack building these documents. You’d have to think of the musicians as being just like those first people who put blink tags all over their web sites as they learned HTML. They’d be lost, but might be motivated.

That’s a pretty huge deal, since without musicians making sites there won’t be too much going on, but I do have some insight into the newer musicians. Younger guys have no problem using technology to advance their music to others. Older musicians, even just the ones that are my age, seem to be keen to learn the stuff, but lack the skills many times.

I don’t have any answers for how musicians would produce, it’s the classic problem when you start working on something new. There’s so much I’d have to make this work that the most I might do is write about it and give other people ideas. For example, maybe the music production suites like Reason would step up to fill this niche?

Another problem I see is obviously the labels and their army of lawyers. They have successfully sued similar projects as being “mechanisms for helping people find pirated music”. Can you imagine if book publishers sued Tim B.L. to crush the WWW because it could help people find pirated works? Actually, that might have been great for them.

I think it’d be the kind of thing I’d start and just prove it out, and try to keep low key before it was noticed

I Leave You With This

I love this quote on HN by gandalf (man I bet that guy never gets laid):

“So Zed wants to reinvent the wheel, with the help of RSS, to get all the clueless people, clods, and scammers out of the way so that people will discover his own music and the music of other tortured artists.” — gandalf (HN)

It’s funny because I really didn’t say any of that, but like I said, through some people’s lenses I’m whatever they hate.

Another great quote to summarize the above, is something BB King said at a concert in Arizona:

“Playing blues back then was like being black twice.” — BB King

I guess doing code and playing music going to be a double whammy. To techies I’ll be a flake artist, and to artists I’ll be a stuffy nerd. Damn!

Why bother then? If you know people who play guitar (especially guitar) they are complete tech junkies. Myself and many of my friends have blown tons of money on pedals and amps and just random tech. Musicians are totally technical nerds, but they also produce art. When a nerd hears the word art they scoff and say, “Right! Well when you’re done picking apples on your hippie commune you can find me doing calculus!” The truth however is that making music and making code are very similar, and that musicians and geeks are more alike than different, so doing work that combines the two makes me pretty damn happy.

Well, except that musicians get laid like crazy. Even if they suck.